臺灣博碩士論文加值系統

跨交易演算法是一個比較新的研究探討領域，雖然第一篇的文獻來自於1998年，卻比較少有文獻針對此處演算法進行探討，截至目前為止，僅只有二篇文獻針對跨交易資料採礦演算法的研究。以資料採礦的光譜而言，其介於關連式法則與順序型樣之間，且比較偏向順序型樣。由於跨交易演算法可以處理二維的資料屬性，所以在挖掘使用者所感興趣的型樣或規則時，其所耗費的時間與空間複雜度都比傳統的關連式演算法來的高。

我們針對已經發表的兩篇文獻，探究其採用的方法及其所面臨的問題，提出改善效能的解決方案。然後透過分析跨交易演算法的資料特性，提出幾個重要的定理，並實作出效能比較表驗證我們的效能貢獻超過之前的方法。此外，更實際應用到現實的經濟領域，並找到一些有趣的規則。

Inter-transaction mining is a new research area in the domain of data mining. Although the first inter-transaction mining paper was published in 1998, there were very few new algorithms being proposed. As we know now, only two papers [5,7] presented the models and algorithms of intra-transaction mining. On the data mining spectrum, intra-transaction mining is treated between association rule and sequential pattern mining, where it is closer to the sequential pattern. The intra-transaction mining can process two dimensional data such that the required space and time are much more than those of the association rule mining.

In this research we study the effective approaches of mining inter-transactions and focus on improving the solutions presented in [5,7]. We have provided efficient algorithms along with some important lemmas by analyzing the features of inter-transaction. Various experiments were performed to verify that our approach is better than existing methods in [5,7]. Moreover, we applied inter-transaction mining to the financial domain and found some interesting rules from the stock market datasets.

中文摘要 ii
英文摘要 iii
誌 謝 iv
目 錄 v
圖 目 錄 vi
表 目 錄 vii
第一章 導論 8
1.1 簡介 8
1.2 資料採礦 11
第二章 文獻探討 14
第三章 我們提出的方法 17
3.1 問題描述 17
3.2 我們的演算法 21
3.2.1 交易內資料採礦 22
3.2.2 跨交易資料採礦 24
3.3. 效能評估 28
3.4. 結論 31
第四章 實際應用 32
4.1 前言 32
4.2 股市簡介 32
4.3 我們的方法 34
4.3.1 資料來源及屬性 35
4.3.2 股票的選擇 35
4.3.3 資料的前處理 36
4.4.1 定義 38
4.5未來應用 43
第五章 結論及未來研究 44
5.1 結論 44
5.2 未來研究 45
參考文獻 46

[1] Rekesh Agrawal, Ramakrishman Srikant (1995), “Mining Sequential Pattern” Proceedings of 1995 International Conference on Data Engineering, pp. 3–14.
[2]R. Agrawal and R. Srikant. “Fast algorithms for mining association rules.” In VLDB''94, pp. 487-499.
[3]Yen-Liang Chen, Mei-Ching Chiang, Ming-Tat Ko, “Discovering time-interval sequential patterns in sequence databases”, Expert Systems with Applications 25 (2003) 343–354
[4]Jiawei Han, Jian Pei, Yiwen Yin, “Mining Frequent Patterns without Candidate Generation”, 2000 ACM SIGMOD Intl. Conference on Management of Data
[5]Anthony K.H. Tung, Hongjun Lu, Jiawei Han, and Ling Feng, “Efficient Mining of Inter-transaction Association Rules,” IEEE Transactions on Knowledge and Data Engineering, Vol. 15, No.1, Jan/Feb 2003.
[6]Daniel T. Larose, “Discovering Knowledge In Data, An Introduction to Data Mining”
[7]H. Lu, J. Han, and L. Feng, “Stock Movement and n-Dimensional Inter-transaction Association Rules,” Proc. 1998 SIGMOD Workshop Research Issues on Data Mining and Knowledge Discovery, Vol. 12, pp. 1-7, June 1998.
[8]MING-YEN LIN, SUH-YIN LEE “Fast Discovery of Sequential Patterns through Memory Indexing and Database Partitioning”, JOURNAL OF INFO"RMATION SCIENCE AND ENGINEERING 21, 109-128 (2005)
[9]B. Wuthrich, V. Cho, S. Leung, D. Permunetilleke, K. Sankaran, J. Zhang, W. Lam,“Daily Stock Market Forecast from Textual Web Data”, IEEE International Conference on Systems, Man, and Cybernetics, Volume: 3, Page(s): 2720 –2725, 1998.
[10]Chung-Ching Yu, Yen-Liang Chen, “Mining Sequential Patterns from Multidimensional Sequence Data”, IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, VOL. 17, NO. 1, JANUARY 2005
[11]Show-Jane Yen, Arbee L.P. Chen, Member, IEEE, “A Graph-Based Approach for Discovering Various Types of Association Rules”, IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, VOL. 13, NO. 5, SEPTEMBER/OCTOBER 2001
[12]S.Y. Zeng, C.H. Hsueh, S.D. Lee, “Using SOM to Study the Decision Making on Stock Investment,” (in Chinese) Proceedings of the Seventh Conference on Artificial Intelligence and Applications (TAAI2002).